Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4763073 | Chemical Engineering Journal | 2017 | 54 Pages |
Abstract
Particle clusters are important mesoscale flow structures in gas-solid circulating fluidised beds (CFBs). An electrostatic sensing system and two accelerometers are installed on the riser of a CFB test rig to collect signals simultaneously. Cross correlation, Hilbert-Huang transform (HHT), V-statistic analysis, and wavelet transform are applied for signal identification and cluster characterisation near the wall. Solids velocities are obtained through cross correlation. Non-stationary and non-linear characteristics are distinctly exhibited in the Hilbert spectra of the electrostatic and vibration signals, and the cluster dynamic behaviours are represented by the energy distributions of the signal intrinsic mode functions (IMFs). The cycle feature and main cycle frequency of cluster motion are characterised through V-statistic analysis of the vibration signals. Consistent characteristic information about particle clusters is extracted from the electrostatic and vibration signals. Furthermore, a cluster identification criterion for electrostatic signals is proposed, including a fixed and a wavelet dynamic thresholds, based on which the cluster time fraction, average cluster duration time, cluster frequency, and average cluster vertical size are quantified. Especially, the cluster frequency obtained from this criterion agrees well with that from the aforementioned V-statistic analysis. Results from this work provide a new non-intrusive approach to the characterisation of cluster dynamic behaviours and their effects on the flow field.
Related Topics
Physical Sciences and Engineering
Chemical Engineering
Chemical Engineering (General)
Authors
Jingyuan Sun, Yong Yan,